2 persons A & B, each have a happiness metric, determined by their brain. It may be influenced by many factors, such as their probability to survive, reproduce or derivatives of these probabilities. Regardless, they have a happiness metric with current value.
Person A can perform an action that affects the happiness value of person B, that’s a more formal way to say that A is hurting B.
Let’s examine 2 test cases:
Parents of a baby ask the grandmother to babysit the baby. The parents are on a date in a restaurant, as it’s valentine day.
20:00 granny puts the baby to sleep. His happiness goes down to 4.
20:01 Baby cries. Granny’s happiness goes down to 4 as well.
20:03 granny puts baby to play again in the living room. Baby’s happiness back to 9.
20:40 granny calls mother telling her the baby won’t go to sleep. Mother’s happiness down to 5.
21:00 baby won’t go to sleep. Granny calls mother again to complain & ask her to come pick the baby. Mother’s happiness down to 3
21:05 father asks why should they return to pick the baby? Baby is happy, father is happy. Mother says she’s not happy because her mother yelled at her asking her to return. Father happiness down to 6
So who was hurting who here?
Baby -> granny -> mother -> father
We can see that persons being hurt tend to hurt others, especially if they believe these others have hurt them.
This test case also suggests that we have a problem with our definition of hurting (as granny was putting the baby to sleep for his benefit, thinking of his future & not current happiness). To fix our definition, we can add time: lower someone’s current or future happiness metric.
BTW, hurting oneself is usually just hurting your future self, seen as someone else.
a man waits in line for a movie theater cashier. A woman walks straight to the cashier & asks a question about current movies. Than she buys a ticket. The man’s happiness gets down to 7. The man tells her that there’s a line, & she says she thought he was just standing there. The man laughs with the other ppl saying that it’s indeed logical to think that they are just standing in a line there, unrelated to the cashier. Woman’s happiness goes down to 6, because the ppl laughed at her. Man’s happiness up to 8, because he had hurt the woman back. Woman says something nasty about the man to the cashier guy.
Is this test case suggesting that the effect on happiness is an effect of something else, e.g., cutting someone in the queue, i.e., taking something from him without his consent. So, the effect on happiness is an indication of something else which is the real meaning of hurting.
Can 2 non-sentient beings be said to hurt one another? If so, we need to generalize happiness to some other measure.
I’d like to think of it as software instances, that interact with each other. They are programmed to pursue some goals, and have a measure for how much they are getting closer to their goals, in the form of a rewarding feeling. But as they are interacting, they may be hampering each other’s pursue of their goals, eg, by fighting on limited resources. This is indicated by a decrease in the measure. The sequel interactions usually escalate the flow of effects on the measure, among the interacting instances & also other ones getting involved.
This requires better modeling.